So it seems that their definition of 100% means 1 excited state per incoming photon, and then they use a material that converts a single high energy excited state produced by one photon into 2 half as energetic excited states...but then they apply the definition that just counts any excited state per incoming photon to juice their numbers.
So more like 65% energy conversion efficiency at best.
How I wish more people would remember this! What extraordinary interpolation tools we have! Truly marvelous and worth celebrating, but not worth a damn for extrapolation. Such a torrent of well-formed advertising trying to convince experts that a universal and reliable extrapolator has been demonstrated!
The reason this is important is not because the molecule has any practical utility, either now or in the future. Most of the reactions it participates in will probably be various decompositions of itself into other. The real importance comes from what this molecule can teach us about the laws of nature--why Si is the way it is, so distinct from carbon. Specifically, how do the valence electrons of Si manage in a 5 member ring? What does that wave-function look like, and does it agree with our predictions? Do we understand this corner of the universe as well as we think we do? Orbitals are tricky things to compute, so we need difficult to construct molecules to test our calculations.
Huh. That's an interesting possible metric. How many competing tendencies in a space? It a good question and one that's been asked before.
I wonder how machine learnability compares to other measures of chaotic structure, like multi-fractal approaches etc. I wouldn't be that surprised if it's accidentally the same or quite similar to some of the existing metrics.
So more like 65% energy conversion efficiency at best.